Gait variability has been traditionally defined as fluctuations in gait characteristics such as step length, stance time or step width and is most often measured by standard deviation (SD) or coefficient of variation (CV). Gait variability has been shown to be a better indicator of future falls than mean spatial and temporal gait characteristics in persons who walk at near normal speed(8). Increased variability in several gait indicators, including stride time, step length, and stance time, is associated with falling, while the association of step width variability with falls has a more inconsistent relationship's(5) An increase in absolute step width is thought to lead to greater stability. Some investigators have foimd step width to be reduced in fallers(4) while others have found a nonlinear association, with both too little and too much step width variability related to falls(8)